Abstract: A physical model of continuous induction hardening process for S45C steel step shaft was established, which was reasonably divided into areas and loaded with corresponding physical properties and process parameters. The hardening depth distribution of the hardened layer was predicted using the radial- axis temperature distribution at the end of the heating phase, the Maynier model, the Carsi correction model and the steel CCT curve fitting. The parameters of continuous induction hardening process were optimized by jointly analyzing the simulation results of each method. The results show that the depths of 100%, 50% and 0% martensite in the induction hardened layer of the tested key point b are 1.31, 1.49 and 2.97 mm with errors of -12.67%, -13.87% and -1.00%, respectively, as well as the key point e are 1.44, 2.02 and 2.54 mm with errors of -4.00%, -3.38% and -18.06%, which are in good agreement with the test results. The quasi-optimized induction hardening process parameters were discussed by changing the induction hardening process parameter and adjusting the physical model. The effect of heat exchange coefficient h change on the induction hardening layer was preliminarily investigated.

Key words: continuous induction hardening, stepped shaft, hardened layer, prediction methods, process optimization